Use cases
Each scenario answers three questions: what becomes possible, how Strands helps, and where it stands today. Demonstrated means shown working on our test systems. In development means being built. Concept means the architecture is designed for it, but it has not been built or proven.
A hospital can keep your records off-site without being able to read them. Your doctor reads them because you granted it, even while your own computer is off. Change your mind and the access stops.
How Strands helps: your information lives in sealed envelopes you hold the keys to; custodians store copies they cannot open.
Show you're over 18 without revealing your birthday. Answer a tax office's questions without opening your whole file. Organisations get the information they need, rather than the information they want.
How Strands helps: a request gets one approved answer, scoped, time-limited and auditable. No central directory, no universal identifier, nothing to trawl.
Keep your medical self, your driving record and your gaming persona apart, with no way to link them unless you choose. The family's machines work as one; each person stays in charge of their own.
How Strands helps: separate envelopes per part of your life, each with its own grants.
Microscope or spectrometer, processing, analysis, recording and visualisation connected as reusable pieces, independent of which vendor made which part.
How Strands helps: each step is a node with a declared interface. Instruments need adapters, which are not built yet.
Imagine a collaboration on the scale of ITER or CERN: each institution runs its own experiments and builds its own analysis, then shares a chosen piece or dataset with partners, not its whole infrastructure.
How Strands helps: bilateral, revocable grants of specific capabilities. Federation is not a merger.
Distributed instruments preprocess locally, send what matters for remote analysis, and every result keeps its provenance and an accountable home.
How Strands helps: the same pipeline runs whether its pieces share a room or a continent. Cross-site fabrics are in development.
Researchers work with exactly the data they've been granted, inside a defined workflow, and every use is recorded.
How Strands helps: scoped authority and provenance. This does not by itself make a project compliant with any regulation.
Collaborate across institutions without anyone assuming central ownership of every dataset.
How Strands helps: each party remains the authority over its own data and grants only what the collaboration needs.
Buildings, factories, aircraft and cities modelled from live telemetry, simulations, records and views working together.
How Strands helps: each source is a capability in the fabric, composed into one living model.
Every workstation, server room and office runs the same system and is managed from a desk. Software comes only from publishers you trust, and a machine can be cut off the moment you stop trusting it.
How Strands helps: one system image in different roles, signed software and immediate trust withdrawal. Granting a partner one specific service is in development.
Several people with their own working environments, sharing the infrastructure behind them, supplied and maintained remotely.
How Strands helps: a desktop is an assembly of capabilities, not a box, so it can draw on resources elsewhere.
Plug in a new inspection camera or analysis step without rewriting the whole system.
How Strands helps: new capabilities join the fabric through declared interfaces. Safety-critical controls stay local and protected.
Cameras, perception, planning and supervision cooperate as services across machines.
How Strands helps: shared capabilities for perception and planning. Real-time actuation and emergency stops must never depend on best-effort network messages.
Deferrable work runs where power, cooling, latency and cost are most favourable.
How Strands helps: work is placed by capability rather than tied to one machine. Scheduling for energy is not built.
Control rooms with many screens and inputs, and edge sites that respond promptly and report only what matters.
How Strands helps: the system is event-driven: work happens when something happens.
A cooperative owns shared compute and storage; members keep their own environments, under rules everyone can see.
How Strands helps: shared infrastructure with individual authority and explicit, inspectable grants.
Services spread across machines degrade gracefully and recover when a machine or link comes back.
How Strands helps: custody and recovery are designed in. We don't promise behaviour under network partitions until it's tested.
Local autonomy while out of contact, with synchronisation when communications return.
How Strands helps: a future design target, not a current capability.
The big picture
Every country's air navigation service runs its own systems and answers for its own skies. Yet every flight has to pass smoothly from one to the next, with each side sharing exactly what the handover needs, and nothing it shouldn't.
Independent authorities, each in full control of its own part, working together as one system: that is the shape of problem Strands was designed for. It is a vision, not a product, and nobody is flying on it.
Many parties, each in control of its own part, needing to work as one. Tell us about it.
